Migration guide · 10 minutes
AWS Bedrock → QuickSilver Pro
Unlike the reseller migrations, this one swaps SDKs: you drop the AWS SDK + SigV4 signing and use the plain OpenAI SDK with a Bearer key. QuickSilver Pro's DeepSeek V4 Pro is the closest reasoning fit at $0.70 / $2.10 per 1M, plus you get the whole V4 wave, Qwen 3.6, and Kimi K2.6 that Bedrock doesn't carry. For the full side-by-side analysis, see /vs/bedrock.
- Code changes
- SDK · auth · model IDs
- Model IDs to rename
- 2 mappings
- Then
- test your core flows end-to-end
The steps
Sign up at quicksilverpro.io/dashboard. First top-up bonus: we match your first top-up 100%, up to $50 — added to your balance when you top up again (any amount from $5).
Bedrock requires the AWS SDK (boto3 / aws-sdk) and SigV4 request signing. QuickSilver Pro serves the OpenAI chat-completions shape directly, so you drop that plumbing entirely and use the official openai client with a Bearer API key.
Bedrock's Converse API maps cleanly to OpenAI chat completions — streaming, tool calling, and structured output behavior carry over. See the full before/after below.
Bedrock uses region-qualified inference-profile IDs. QuickSilver Pro uses short model names. Bedrock does not carry the DeepSeek V4 wave, Qwen 3.6, or Kimi K2.6 — those become newly available on QuickSilver Pro; DeepSeek V4 Pro is the closest reasoning replacement for Bedrock's DeepSeek offering.
AWS Bedrock QuickSilver Pro Current Bedrock DeepSeek model deepseek- v4- pro (not on Bedrock) deepseek- v4- flash Run one representative request for each feature you use — chat, streaming, tool / function calling, and json_schema strict mode. Apart from the known difference below, any behavioral difference is a bug — report it.
One known difference: json_schema strict mode is model-dependent. The Claude models do not support it, so a schema there is advisory and the enforced path is a tool with
strict: true— structured output.
Full before/after
import boto3, json
client = boto3.client("bedrock-runtime", region_name="us-east-1")
resp = client.converse(
modelId=os.environ["BEDROCK_DEEPSEEK_MODEL_ID"],
messages=[{"role": "user", "content": [{"text": "Hi"}]}],
)
text = resp["output"]["message"]["content"][0]["text"]import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.quicksilverpro.io/v1",
api_key=os.environ["QSP_KEY"],
)
r = client.chat.completions.create(
model="deepseek-v4-pro",
messages=[{"role": "user", "content": "Hi"}],
)
text = r.choices[0].message.contentimport boto3, json
client = boto3.client("bedrock-runtime", region_name="us-east-1")
resp = client.converse(
modelId=os.environ["BEDROCK_DEEPSEEK_MODEL_ID"],
messages=[{"role": "user", "content": [{"text": "Hi"}]}],
)
text = resp["output"]["message"]["content"][0]["text"]What you'll pay after switching
Per 1M tokens, input / output. QuickSilver Pro rates vs AWS Bedrock's published per-token pricing.
| Model | QuickSilver Pro | AWS Bedrock | Savings |
|---|---|---|---|
| DeepSeek V4 Pro | $0.70 / $2.10 | $0.55 / $2.19 | ~parity |
| DeepSeek V4 Flash | $0.06 / $0.18 | — / — | not on Bedrock |
AWS Bedrock prices: no check date
Common migration pitfalls
You lose AWS-native plumbing
IAM-scoped access, VPC endpoints, Bedrock Guardrails, and Knowledge Bases have no QuickSilver Pro equivalent. If any of these are load-bearing for your compliance story, keep that traffic on Bedrock — or move only the cost-sensitive subset.
Auth model changes from SigV4 to Bearer
Bedrock signs every request with SigV4 derived from AWS credentials. QuickSilver Pro uses a static Bearer API key. Update your secrets management accordingly and treat the key like any other credential.
Not every Bedrock model exists on QuickSilver Pro
QuickSilver Pro supports selected Claude models alongside DeepSeek, Qwen, Kimi, GLM, and other families, but it does not mirror Bedrock's full catalog or accept Bedrock-specific model IDs. Verify each workload against the QuickSilver Pro models reference, and keep unsupported Nova, Llama, Mistral, or model-revision dependencies on Bedrock.
Region pinning differs
Bedrock inference is region-locked to your AWS region. QuickSilver Pro runs on dedicated bare-metal in Europe with US edge. If data residency is a contractual requirement, evaluate this before moving.
Migrating from AWS Bedrock — FAQ
Compare catalog prices, context windows and capabilities. Choose models using your own workload. Claude Fable 5.2: $2.00 input / $10.00 output per 1M tokens. Compare current catalog rates.
If IAM / VPC endpoints / Guardrails / Knowledge Bases / BYO-VPC are load-bearing for your compliance story, or if you need Claude / Nova / closed models in the same provider. The Bedrock premium pays for AWS-native integration; QuickSilver Pro is for teams who don't need that and would rather not pay the markup.
Drop the AWS SDK / SigV4 plumbing and use the openai SDK directly: base_url="https://api.quicksilverpro.io/v1", api_key="$QSP_KEY". Bedrock's Converse API is broadly equivalent in capability to OpenAI chat completions; QSP serves the OpenAI shape directly so any OpenAI SDK works unchanged.
Not natively. QSP is focused on raw chat-completions inference at a narrow surface and predictable price — not a wider LLM platform. For Guardrails-equivalent content moderation, drop in any open-source moderation library against the chat output. For RAG, our 1M-context V4 Pro and 262K-context Qwen 3.6 let you skip the vector store entirely on small-to-mid corpora.
Start saving in 10 minutes
First top-up matched 100%, up to $50 in bonus credits, added on your next top-up. Swap the AWS SDK and SigV4 signing for the OpenAI SDK — then it's a base URL and a Bearer key.